ISSN 0972-978X 

  About COAA








Post-Operative Pyrexia Following Lower Limb Arthroplasty A Comparative Clinical Study

 Eitan Melamed *, Dror Robinson, Moshe Weisbrot*, Anatoli Pirotzki*, Avraham Garti*,

* Department of Orthopaedic Surgery, Rabin Medical Center , Golda Campus,  7 Keren Kayemet Street , Petach-Tikva, 49372, Israel .

Address for Correspondence:

Eitan Melamed, MD,
Department of Orthopaedic Surgery, Rabin Medical Center , Golda Campus,
7 Keren Kayemet Street , Petach-Tikva, 49372, Israel . 
Tel      : 972-3-9372236, Fax- 972-3-9372501.



Introduction: Post-operative pyrexia is a common cause of concern for both surgeon and patient.  The current study was performed in order to assess the significance of this common symptom and the need for prolonged hospitalization and diagnostic work-up in such patients.
Methods: We conducted a retrospective analysis of 77 consecutive joint arthoplasties performed at our institute.  In all cases, office notes indicate lack of a surgical site infection for a minimum of two years.  Demographic parameters, type of anesthesia, blood loss and transfusion requirements as well as leukocyte counts, urine cultures and chest radiographs were correlated to temperature measurements.  A febrile response was considered oral temperature of more than 38 degrees Celsius.
Results: 58% of patients had a febrile response. This most commonly occurred on the first and second post-operative days.  The febrile response was not correlated to amount of blood loss, homologous transfusion, positive urine cultures or leukocytosis.  However, general anesthesia appears to be associated with a febrile response perhaps due to pulmonary complications. 
Conclusions: A febrile response in the peri-operative arthroplasty patient does not indicate a surgical site infection.  It might be related to the use of general anesthesia with some attendant pulmonary damage.  Hospital discharge should not be delayed due to peri-operative pyrexia.

J.Orthopaedics 2007;4(3)e15

arthroplasty, pyrexia, inflammation, knee, hip, peri-operative.


Postoperative pyrexia in patients undergoing total joint arthroplasty procedure is often alarming for both staff and patients.  Current surgical practice is to delay discharge of such patients until they are afebrile for at least two days.  This is a wasteful practice, which is especially difficult to accept in today's cost-conscious day and age.  The current study was conducted in order to assess the significance of post-operative fever.  Furthermore, an attempt to define the variables responsible for pyrexia has been made.  Based on results of this study, a treatment algorithm for the patient with peri-operative post-arthroplasty fever is suggested.

Material and Methods :

The medical records of 77 consecutive patients were evaluated.  Pyrexia was defined as a temperature higher than 38 degrees Celsius.  The variables of concomitant diseases, pre-operative and post-operative hemoglobin, amount of drained blood, pre-operative and post-operative white blood cell count, amount of transfused blood, results of urine cultures and results of chest radiographs were evaluated.  Daily Q 4 hour measurements of oral temperature were made using a digital thermometer accurate to one tenth of a degree Celsius.  The day of surgery started in the recovery room and continued until 6:00 AM on the following day.

Postoperative Day one began at 6:00 AM on the day after surgery and continued for 24 hours. Subsequent postoperative days began and ended at 6:00 AM in a repeating sequence. All patients in this series received subcutaneous enoxaparin 40mg for six weeks.  All received peri-operative antibiotics (cefonicide a second generation cephalosporin) as a single pre-operative dose.  The use of incentive spirometers was encouraged in all patients.  All patients began formal physical therapy and gait training on postoperative day one.  Patients were hospitalized for 4 to 8 days (average hospitalization was 5.23 days).

The procedures were performed in one university hospital during an 8 months interval.  All were primary replacement procedures. 

All patients had hospital records documenting a minimum of 2 years post-operative follow-up without evidence of prosthetic joint infection. No attempt was made to define clinical outcome parameters other than the absence of peri-operative infection.

No surgery was done if the patient had known pre-operative intercurrent infection. Patients with immuno-compromising disease or receiving immuno-suppressive drugs were excluded (diabetic patients were not excluded).  

Demographic characteristics of the patients are recorded in Table 1.  To define variables that may have etiologic significance relative to postoperative elevated temperatures, multiple subgroups were examined within the combined populations of patients who had total knee or hip replacements. Comparative groups included:

Patients who had a total knee replacement versus patients who had a total hip replacement; patients who had a general anesthetic versus patients who had epidural anesthesia; patients who received no blood transfusion (at our institute only allogeneic blood is transfused) versus patient who received blood; patients with positive results of a urine culture versus patients with negative results of a urine culture; as well as patients with an abnormal chest radiograph versus patients without an abnormal chest radiograph.

This was a retrospective patient data analysis and thus informed consent had not been obtained.  The institutional Ethical Review Board approved the study.

Statistical Analysis was performed using the Analyze-It add-in to Microsoft Excel XP (Analyze-It Ltd., version 1.71, 2003).

Results :

58% of patients had oral temperatures higher than 38 degrees Celsius measured during their hospitalization.  Most commonly this occurred during the second postoperative day.  Maximal temperature is similar in total knee arthroplasties (38.1+0.5) as compared to total hip arthroplasty (38.00.5, t-test, p>0.6).  Average daily temperature was also similar in both of these groups.

Leukocytosis was not associated with elevated temperature (Pearson correlation, r-value 0.06, p>0.7) or with abnormal chest radiographs (1-way between subjects ANOVA, F-value=0.6, p>0.6) or with positive urine cultures radiographs (1-way between subjects ANOVA, F-value=0.3, p>0.7).  There was no correlation between temperature and amount of blood drained from the wound (Pearson correlation, r-value 0.04, p>0.7). 

Type of anesthesia influences the maximal temperature recorded.  Patients undergoing general anesthesia had significantly higher maximal temperature (38.30.5) than those undergoing spinal anesthesia (38.10.4, t-test, p<0.05).  General anesthesia was associated as well with a significantly larger decline in hemoglobin (9.31.3 gram per deciliter) as compared to spinal anesthesia (101.2 gram per deciliter, t-test, p<0.05).  Hemoglobin values remained significantly lower at discharge in patients undergoing general anesthesia  (10.10.9 gram per deciliter) as compared to patients undergoing spinal anesthesia (10.61, t-test, p<0.03). 

Fifteen patients (19%) were transfused.  Their maximal as well as their average temperature was similar to that of those patients who were not transfused (t-test, p>0.12).  Transfusion did not appear to be associated with leukocytosis (t-test, p>0.09).  

Forty patients (51%) had urine cultures performed due to temperature elevation.  There was a significant difference in temperature between those patients in whom urine cultures were not performed (maximal temperature 380.4) then in these patients in whom cultures were performed (38.30.5).  However there was no significant difference between those patients in whom cultures were performed whatever the culture results (ANOVA, F-value 2.9, p<0.05, Dunnet 95% confidence interval, difference was 0.03).   Chest radiographs were performed in 26 patients (36%) due to clinical findings during physical examination.  A positive chest radiograph was significantly associated with elevated temperature (ANOVA, F-value 11.6, p<0.001, the difference between patients that had an abnormal chest radiograph and those that did not have a radiograph was significant, Dunnet 95% confidence).  Six patients had blood cultures due to elevated temperature.  All blood cultures were negative.

The type of anesthesia appears to predispose to pyrexia, i.e. general anesthesia is related to post-operative pyrexia (1-way between subjects ANOVA, F-statistic=3.8, p<0.05), perhaps due to pulmonary complications as evidenced by abnormal chest radiographs (Kruskal-Wallis ANOVA, KW Statistic=24.6, p<0.001). 

Urine cultures are quite often positive in the post-arthroplasty patient, perhaps due to their average age and debilitation.  However, no statistically significant correlation is found between the presence of positive urine culture and postoperative pyrexia (Kruskal-Wallis ANOVA, KW statistic 3.5, p>0.05).   

Discussion :

The study appears to indicate that pyrexia occurs in the majority of patients following lower limb arthroplasty.  This finding is probably attributable to an inflammatory response of the body to the surgical insult, which occurs fairly uniformly after hip operations [1]. There does not appear to be any correlation between elevated temperatures after knee or hip arthroplasty and infection of the surgical site.  According to the current study, there is no correlation between positive urine cultures and post-operative pyrexia.  This finding concurs with that of Shaw et al. who have found no relationship between pyrexia and a positive urine culture [2]. In another study, patients with positive urine cultures uniformly did not have pyrexia [3]. However, a positive chest radiograph appears to be correlated with elevated temperatures.  This probably means that patients with elevated temperatures and abnormal pulmonary physical findings are likely to have an abnormal finding on chest radiographs.  However, many of these findings are non-specific and do not necessarily indicate a cause and effect relationship to the pyrexia.

There was no relation between pyrexia and bleeding-related parameters (including amount of blood lost, blood transfusion or pre-hospitalization or pre-discharge hemoglobin, ANOVA).  Kennedy et al [3] reported a weak correlation that did not reach statistical significance between perioperative blood loss and pyrexia in a series of 92 consecutive TKA.  In our series no such correlation was observed. 

This finding is similar to what has been reported by Kennedy et al [3] but contrasts with the observation reported by Guinn and colleagues that urine analysis might explain some cases of post-operative pyrexia [4].  

What is the reason for the pyrexia?  The relationship to general anesthesia might indicate that pulmonary dysfunction or perhaps infective agents introduced into the pulmonary system are involved.  However, it has been shown that an intense inflammatory response occurs after lower limb arthroplasties due to cytokine secretion.  A major cause might be interleukin-6, whose level is increased following lower limb arthroplasty, and has been shown to be responsible for febrile reactions in the post-arthroplasty patient [5].   Another factor might be the formation of platelet-leukocytes complexes concomitantly with interleukin-6 secretion, which occurs after arthroplasty, lasts for about two weeks and is responsible for an inflammatory response as well as activation of the thrombotic mechanism [6].  

The current study appears to indicate that post-operative pyrexia after lower limb arthroplasty is not a cause for concern from the perspective of possible surgical site infection, but rather a normal expected finding.  There does not appear to be any reason to delay discharge of patients due to pyrexia without localizing physical findings.  However, the inflammatory response accompanying arthroplasty is important as increased interleukin-6 levels are associated with delayed functional rehabilitation post-operatively [7]. Thus, it appears of importance to include a mechanism for decreasing the inflammatory response post-arthroplasty, in rehabilitation protocol of arthroplasty patients.


Total Hip Arthroplasty    N=47

Total Knee Arthoplasty    N=30

Statistical Significance



Osteoarthritis 35

Aseptic Necrosis 10

Rheumatoid arthritis 2

Osteoarthritis 25

Aseptic Necrosis 2

Rheumatoid arthritis 3

Unsuitable numbers to compare




t-test = 0.2, 2-tailed p>0.8

Diabetes mellitus

11 patients

8 patients

Mann-Whitney U statistic 713, p>0.9

Ischemic Heart Disease

8 patients

6 patients

Mann-Whitney U statistic 711, p>0.9


13 patients

9 patients

Mann-Whitney U statistic 723, p>0.8  

Febrile response

28 patients

17 patients

Mann-Whitney U statistic 684, p>0.8

Table 1: Demographic Characteristics of Patients included in the Study

A cut-off value of p<0.05 was considered as significant.  T-test for independent samples (two-tailed) was used for contiuous variables.  Mann-Whitney test or Kruskal Wallis were used for nominal variables.

Reference :

  1. Okafor B, MacLellan G: Postoperative changes of erythrocyte sedimentation rate, plasma viscosity and C-reactive protein levels after hip surgery. Acta Orthop Belg 1998, 64:52-56.

  2. Shaw JA, Chung R: Febrile response after knee and hip arthroplasty. Clin Orthop Relat Res 1999, 367:181-189.

  3. Kennedy JG, Rodgers WB, Zurakowski D, Sullivan R, Griffin D, Beardsley W, et al: Pyrexia after total knee replacement. A cause for concern? Am J Orthop 1997, 26:549-552, 554.

  4. Guinn S, Castro FP, Jr, Garcia R, Barrack RL: Fever following total knee arthroplasty. Am J Knee Surg 1999, 12:161-164.

  5. Handel M, Winkler J, Hornlein RF, Northoff H, Heeg P, Teschner M, et al: Increased interleukin-6 in collected drainage blood after total knee arthroplasty: an association with febrile reactions during retransfusion. Acta Orthop Scand  2001, 72:270-272.

  6. Bunescu A, Widman J, Lenkei R, Menyes P, Levin K, Egberg N: Increases in circulating levels of monocyte-platelet and neutrophil-platelet complexes following hip arthroplasty. Clin Sci (Lond) 2002, 102:279-286.

  7. Hall GM, Peerbhoy D, Shenkin A, Parker CJ, Salmon P: Relationship of the functional recovery after hip arthroplasty to the neuroendocrine and inflammatory responses. Br J Anaesth 2001, 87:537-542.


This is a peer reviewed paper 

Please cite as : Eitan Melamed : Post-Operative Pyrexia Following Lower Limb Arthroplasty A Comparative Clinical Study

J.Orthopaedics 2007;4(3)e15





Arthrocon 2011

Refresher Course in Hip Arthroplasty

13th March,  2011

At Malabar Palace,
Calicut, Kerala, India

Download Registration Form

For Details
Dr Anwar Marthya,
Ph:+91 9961303044



Powered by



Copyright of articles belongs to the respective authors unless otherwise specified.Verbatim copying, redistribution and storage of this article permitted provided no restrictions are imposed on the access and a hyperlink to the original article in Journal of Orthopaedics maintained. All opinion stated are exclusively that of the author(s).
Journal of Orthopaedics upholds the policy of Open Access to Scientific literature.